Abstract
Background
Candida species are emerging as a potentially pathogenic fungus in patients with broncho-pulmonary diseases. The synergistic growth promoting association of Candida and Mycobacterium tuberculosis has raised increased concern for studying the various Candida spp . and its significance in pulmonary tuberculosis patients during current years.
Aims
This study was undertaken with the objective of discovering the prevalence of co-infection caused by different Candida species in patients with pulmonary tuberculosis.
Method
A total of 75 patients with pulmonary tuberculosis diagnosed by sputum Ziehl-Neelsen staining were included in the study. Candida co-infection was confirmed using the Kahanpaa et al. criteria. Candida species were identified using gram stain morphology, germ tube formation, morphology on cornmeal agar with Tween-80, sugar fermentation tests and HiCrome Candida Agar.
Results
Candida co-infection was observed in 30 (40%) of patients with pulmonary tuberculosis. Candida albicans was the most common isolate observed in 50% of the patients with co-infection, followed by C. tropicalis (20%) and C. glabrata (20%). Candida co-infection was found in 62.5% of female patients, while it was observed in only 29.4% of the male patients (P value 0.0133). Mean ± SD age of the patients with C. glabrata infection was 65.83 ± 3.19, while the mean ± SD age of the patients with other Candida infections was 43.25 ± 20.44 (P value 0.0138).
Conclusion
Many patients with pulmonary tuberculosis have co-infection with Candida spp. The prevalence of non-albicans Candida species is increasing and may be associated with inadequate response to anti-tubercular drugs. C. glabrata infection has a strong association with old age.
Keywords: Candida co-infection, C. glabrata, prevalence, tuberculosis
What this study adds:
Candida is an opportunistic fungal pathogen infecting immunocompromised hosts. Although the synergistic growth promoting association of Candida and Mycobacterium tuberculosis is well documented, sputum isolates of Candida spp . are usually ignored as an innocuous throat commensal.
A significant proportion of isolates were non-albicans Candida species which were also more frequent among tuberculosis patients with persistence of chest symptoms in spite of anti-tubercular treatment of two months or more.
This study implies the need for screening pulmonary tuberculosis patients for Candida co-infection, especially in case of patients with inadequate response to anti-tubercular therapy. Anti-fungal sensitivity test may be valuable in case of non-albicans Candida species.
Background
Candida albicans has emerged as a potentially pathogenic fungus rather than innocuous mucosal commensal in patients with broncho-pulmonary diseases. Although respiratory candidiasis secondary to pulmonary tuberculosis has been reported in the past,1 it has gained more relevance recently due to increased use of broad spectrum antibiotics and immunosuppressive drugs2 and possibly as a result of resurgence of tuberculosis in the background of the HIV epidemic.3 The synergistic growth-promoting association of Candida and M. tuberculosis has also been documented experimentally.4 There is increased concern with studying altered mycotic respiratory flora and its significance in pulmonary tuberculosis patients in current years due to this change in trends.2 Although Candida albicans continues to be the most predominant species in pulmonary candidiasis,2,5-7 several non albicans Candida species are also reported in increasing frequency. Some of them are associated with particular risk factors or groups of patients.5-9
The present study was undertaken with an objective of discovering the prevalence of co-infection caused by different Candida in patients with pulmonary tuberculosis.
Method
This study was carried out in a tertiary care hospital in South India over a period of nine months (from June 2010 to March 2011). Early morning sputum samples were collected on three consecutive days from patients with suspected pulmonary tuberculosis (clinically and radiologically). Sputum samples of 382 patients from both hospital in-patients and out-patients were received within the study period and were processed routinely by gram stain, Ziehl-Neelsen stain, KOH mount, culture in blood agar and McConkey’s agar. In sputum gram stain number of pus cells and epithelial cells per low power field, bacteria, presence or absence of fungal elements were noted.
Out of 382 patients, 75 patients were sputum positive for acid-fast bacilli in Ziehl-Neelsen stain and were further processed by inoculating in Sabouraud’s dextrose agar. Gram stain was done from suspected yeast colonies. Of the 75 patients positive for acid-fast bacilli, only those with budding yeast cells and pseudohyphae along with pus cells in sputum gram stain and heavy growth of Candida with more than 30 colonies on SDA on three occasions were considered to have Candida co-infection, as per Kahanpaa et al. criteria to exclude respiratory or oral colonising flora.10 Candida species were differentiated from other yeasts and were identified up to species level using gram stain morphology, germ tube formation, cornmeal agar with Tween-80 (for demonstration of chlamydospore, blastospores and pseudohyphae), urease test, sugar fermentation tests (glucose, sucrose, lactose and maltose) and HiCrome Candida agar. Candida isolates displaying pseudohyphae with clusters of spherical blastoconidia at the constriction sites and thick walled large round terminal single chlamydospores on cornmeal agar were considered as C. albicans . C. parapsilosis strains were identified by blastoconidia in single or in small groups along branched pseudohyphae which gradually become smaller (Christmas tree appearance). By contrast, the long branching abundant pseudohyphae with narrow sterile apex and ellipsoid blastoconidia of C. tropicalis , clusters of small, budding blastospores without pseudohyphae is very typical of C. glabrata .11 Chrom agar colony morphology, sugar fermentation and germ tube test findings were also correlated with the microscopic findings for species identification.
All findings were entered in MS Excel data sheet and on completion of the study data was statistically analysed in SPSS software version 17.0 by calculation of two tailed P value using chi-square test. All P values < 0.05 were considered as statistically significant.
Results
Out of the total 382 patients with suspected tuberculosis, 75 patients (19.6%) who were positive for acid fast bacilli by Ziehl-Neelsen stain were screened for Candida co-infection. Among the 75 patients with pulmonary tuberculosis, Candida co-infection was observed in 30 (40%) patients. Candida albicans was the most common isolate observed in 50% of the patients with co-infection, followed by C. tropicalis (20%) and C. glabrata (20%) (Table 1). There was no significant difference in the mean age of the patients with and without Candida co-infection (Table 2). There was a significant female preponderance for occurrence of Candida co-infection. Candida co-infection was found in 62.5% female patients, while it was observed in only 29.4% of the male patients (P value 0.0133) (Table 2). Candida co-infection was observed in 45.0% of the out-patients with pulmonary tuberculosis, while it was noticed in only 20.0% of the in-patients. However, this difference was not statistically significant (Table 2).
Table 1 . Distribution of various Candida spp . from patients with tuberculosis.
| Candida species | Frequency | Per cent |
|---|---|---|
| C.albicans | 15 | 50.0 |
| C. tropicalis | 6 | 20.0 |
| C. glabrata | 6 | 20.0 |
| C. parapsilosis | 2 | 6.7 |
| C. krusei | 1 | 3.3 |
| Total | 30 | 100.0 |
Table 2. Comparison of the characteristics of patients with and without Candida co-infection.
| Characteristic | Candida co-infection (n = 30) | No Candida infection (n=45) | P value |
|---|---|---|---|
| Mean ± SD age | 47.77 ± 20.43 | 48.80 ± 13.72 | 0.79 |
| Sex Female (n=24) | 15 (62.5%) | 09 (37.5%) | 0.01 |
| Male (n=51) | 15 (29.4%) | 15 (29.4%) | |
| Category In-patient (n=15) | 03 (20.0%) | 12 (80.0%) | 0.14 |
| Out-patient (n=60) | 27 (45.0%) | 33 (55.0%) |
Mean ± SD age of the patients with C. glabrata infection was 65.83 ± 3.19, while the mean ± SD age of the patients with other Candida infections was 43.25 ± 20.44 (P value 0.0138). C. glabrata infection was found in 40% (6/15) female patients with Candida co-infection, while it was not observed among the 15 male patients with co-infection (P value 0.0169).
Among the 30 patient with Candida co-infection, 19 (63%) had had pulmonary symptoms in spite of anti-tubercular treatment of two months or more. Although, Candida albicans was the most common sputum isolate, in this group of patient non-albicans Candida species were more prevalent (12 out of 19 cases).
Discussion
Tuberculosis is well recognised for its wide range of clinical spectrum, chronicity and sequlae. Respiratory fungal infections are one of the emerging conditions complicating pulmonary tuberculosis. Though several authors have documented Candida species as the most common fungal agent isolated from sputum of pulmonary tuberculosis patients, its significance has always been a matter of controversy due to the fact that up to 32.5% healthy people carry Candida in their throat. This can contaminate the sputum sample during collection.12 To eliminate this problem different approaches have been used. Bronchoscopy samples have lesser chance of becoming contaminated with upper respiratory flora and are preferred to sputum.13 Yet in developing countries such as India bronchoscopy is not always feasible or practical. Some studies have detected better correlation with transbronchial biopsies.13 Jain et al. have compared and subtracted the growth of Candida spp. on plain SDA from mouth rinsed water as a control to that of sputum sample.5 Cases which showed response to specific anti-fungal agents were also considered as Candida infection.14 But most commonly this evaluation is done by criteria suggested by Kahanpaa et al.10 According to this criteria three or more repeated isolations of Candida more than 30 colonies on SDA with pseudomycelial forms in sputum microscopy is more suggestive of infection than colonisation. In this study we have also followed this criteria to detect Candida infection.
However, this is not applicable for C. glabrata which has only yeast forms and some studies failed to find the significance of pseudohyphal forms in respiratory samples.
The prevalence of Candida co-infection of lung ranges between 15-32% in different studies.5, 15-17 With some rare exceptions,15 C. albicans has been reported to be the most predominant isolate from sputum of tuberculosis patients followed by C. tropicalis .2, 5, 16, 17 Wide variation ranging from 45–92% was seen in the prevalence of C. albicans in several Indian studies.2, 5, 16, 17 In the present study we detected 50%, 20%, 20%, 6.7% and 3.3% prevalence of C. albicans, C. tropicalis, C. glabrata, C. parapsilosis, and C. krusei respectively. This result is in keeping with other similar studies.2, 5, 16 Jain et al. reported C. tropicalis (9.1%), C. pseudotropicalis 6.06%, and C. krusei 6.06%.5 Baradkar et al. detected C. tropicalis 3.25%, Candida parapsilosis 3.25%.16 Latha et al. documented Candida tropicalis (19.95%), Candida glabrata (16.54%), C. parapsilopsis (13.14%) and C. krusei (5.10%).2 These variations in percentages are mainly attributed to differences in local prevalence of different species due to different environmental conditions, as well as to the various detection methods employed.5,17 Respiratory specimens like BAL, bronchial wash which precludes oropharyngeal contamination may also affect the result.17 According to Hidalgo et al. colonisation rates of Candida species are equal in males and females.18 However, we observed that Candida co-infection was significantly higher among female patients compared to male patients (P value 0.0133). The relatively high colonisation rates in women could have been responsible for increased risk of Candida co-infection among female patients with pulmonary tuberculosis. We found no statistically significant difference in Candida co-infection between in-patients and out-patients. However, patients with C. glabrata infection were predominantly female (P value 0.0168) as well as older (mean age + SD 65.83 ± 3.19 years) compared to patients with other Candida spp. (mean age + SD 25 ± 20.44 years) (P value 0.0138). Although old age is a known risk factor for C. glabrata, 7 there is no clear evidence to suggest higher susceptibility in females for pulmonary infection by C. glabrata than other Candida species. It is notorious for acquisition of azole resistance (especially fluconazole) as well as cross resistance. It is emerging as a successful nosocomial pathogen in presence of risk factors like immunocompromised or old debilitated host, prolonged hospitalisation, antibiotic use, fluconazole use, hospital acquired exposures from infected patients, hands of healthcare workers and environment in contact with hands.7
A significant increase in the number of cases of HIV infection has been reported in tuberculosis patients in recent years. It is often described as a HIV, tuberculosis co-epidemic.19 However, testing for HIV infection in tuberculosis patients has not emerged as a routine practice in resource limited and developing countries like India. Only 22 (29.3%) out of 75 patients were tested for HIV and found negative. Apart from weight loss and chronic cough, none had symptoms suggestive of HIV infection.
All patients were started on anti-tubercular drugs (ATD) on diagnosis. Among them, 67 new cases received category I regimen of Revised National Tuberculosis Control Programme (RNTCP) of India, comprising of isoniazid, rifampicin, pyrazinamide and ethambutol for two months (intensive phase) followed by isoniazid and rifampicin for four months (continuation phase).20 On follow-up sputum microscopy for acid fast bacilli and mycobacterial culture, no cases remained positive at the end of fifth month indicating treatment failure associated with resistance to ATD. There are insufficient published data comparing treatment and their outcomes in pulmonary tuberculosis versus tuberculosis and Candida co-infection. Although, azole resistance is not uncommon among non-albicans Candida species, oral or parenteral azoles are frequently used for pulmonary fungal infections and are preferred to intravenous amphotericin B.21 In a clinical trial, itraconazole showed significant efficacy in patients with Candida and tuberculosis co-infection.22 The first course resulted in complete or partial inhibition of growth of Candida in 68% of patients and after repeated course 84% of patients recovered completely. Only in 2% of cases, it progressed to disseminated pulmonary tuberculosis in association with azole dependent fungal strain of C. glabrata .
In this study we found a shifting pattern of epidemiology of Candida species from commensal to emerging pathogen. Non-albicans Candida species were isolated in increasing numbers from patients with symptomatic lung disease. There is a significant female preponderance for occurrence of Candida co-infection. C. glabrata was found to be strongly associated with old age and female sex. However, due to a small study population the results may not be representative and require further confirmation.
Conclusion
We found increasing prevalence of non-albicans Candida species in pulmonary tuberculosis patients in our hospital. The majority of these patients had persistence of pulmonary symptoms even after anti-tubercular treatment. Owing to the inherent anti-fungal resistance, non-albicans Candida species are often not amenable to anti-fungal treatment. Therefore, screening of pulmonary tuberculosis patient for Candida co-infection should be routinely practiced along with anti-fungal sensitivity testing for non-albicans Candida isolates.
Footnotes
PEER REVIEW
Not commissioned. Externally peer reviewed.
CONFLICTS OF INTEREST
The authors declare that they have no competing interests.
Please cite this paper as: Kali A, Charles MVP, Joseph NM, Umadevi S, Kumar S, Easow JM. Prevalence of Candida co-infection in patients with pulmonary tuberculosis. AMJ 2013, 6, 8, 387-391 http//dx.doi.org/10.4066/AMJ.2013.1709
References
- 1.Schwartin VM, Skinner CE. Candida in sputum of patients with Tuberculosis. Bact Rev. 1948;11:349–55. [Google Scholar]
- 2.Latha R, Sasikala R, Muruganandam N, Venkatesh Babu R. Study on the shifting patterns of Non Candida albicans Candida in lower respiratory tract infections and evaluation of the CHROMagar in identification of the Candida species. J Microbiol Biotech Res. 2011;1:4–9. [Google Scholar]
- 3.Ochieng W, Wanzala P, Bii C, Oishi I, Ichimura H, Lihana R, Mpoke S, Mwaniki D, Okoth FA. Tuberculosis and oral Candida species surveillance in HIV infected individuals in Northern Kenya, and the implications on tuberculin skin test screening for DOPT-P. East Afr Med J. 2005 Dec;82(12):609–13. doi: 10.4314/eamj.v82i12.9365. [DOI] [PubMed] [Google Scholar]
- 4.Mankiewicz E. Mycobacterium tuberculosis and Candida albicans: a study of growth-promoting factors. Can J Microbiol. 1954 Oct;1(2):85–9. doi: 10.1139/m55-012. [DOI] [PubMed] [Google Scholar]
- 5.Jain SK, Agrawal RL, Sharma DA, Agrawal M. Candida in pulmonary tuberculosis. J Postgrad Med. 1982 Jan;28(1):24–9. [PubMed] [Google Scholar]
- 6.Badiee P, Alborzi A. Susceptibility of clinical Candida species isolates to antifungal agents by E-test, Southern Iran: A five year study. Iran J Microbiol. 2011;3:183–8. [PMC free article] [PubMed] [Google Scholar]
- 7.Fidel PL Jr., Vazquez JA, Sobel JD. Candida glabrata: review of epidemiology, pathogenesis, and clinical disease with comparison to C. albicans. Clin Microbiol Rev. 1999;12:80–96. doi: 10.1128/cmr.12.1.80. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Trofa D, Gacser A, Nosanchuk JD. Candida parapsilosis, an emerging fungal pathogen. Clin Microbiol Rev. 2008;21:606–25. doi: 10.1128/CMR.00013-08. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Reuter CW, Morgan MA, Bange FC, Gunzer F, Eder M, Hertenstein B, Ganser A. Candida kefyr as an emerging pathogen causing nosocomial bloodstream infections in neutropenic leukemia patients. Clin Infect Dis. 2005;41:1365–6. doi: 10.1086/497079. [DOI] [PubMed] [Google Scholar]
- 10.Kahanpaa A. Bronchopulmonary occurrence of fungi in adults, especially according to cultivation material. Acta Pathol Microbiol Scand B Microbiol Immunol. 1972:227, 1–147. [PubMed] [Google Scholar]
- 11.Chander J. 2nd. New Delhi: Mehta Publishers; 2002. Textbook of medical mycology. [Google Scholar]
- 12.Santiwongkarn P, Kachonboon S, Thanyasrisung P, Matangkasombut O. Prevalence of oral Candida carriage in Thai adolescents. J Investig Clin Dent. 2012;3:51–5. doi: 10.1111/j.2041-1626.2011.0089.x. [DOI] [PubMed] [Google Scholar]
- 13.El-Ebiary M, Torres A, Fàbregas N, de la Bellacasa JP, Gonzàlez J, Ramirez J, del Baño D, Hernàndez C, Jiménez de Anta MT. Significance of the isolation of Candida species from respiratory samples in critically ill, non-neutropenic patients: An immediate postmortem histologic study. Am J Respir Crit Care Med. 1997 Aug;156(2 Pt 1):583–590. doi: 10.1164/ajrccm.156.2.9612023. [DOI] [PubMed] [Google Scholar]
- 14.Shome SK, Upreti HB, Singh MM, Pamra SP. Mycoses associated with pulmonary tuberculosis. Ind J Tuberc. 1976;23:64–8. [Google Scholar]
- 15.Naz SA, Tariq P. A study of the trend in prevalence of opportunistic Candidal co-infections among patients of pulmonary tuberculosis. Pak J Bot. 2004;36:857–62. [Google Scholar]
- 16.Baradkar VP, Mathur M, Wanjari K, Kumar S. Candida in pulmonary tuberculosis. Bombay Hospital Journal. 2009;51:52–3. [Google Scholar]
- 17.Khanna BK, Nath P, Ansari AH. A study of mycotic flora of respiratory tract in pulmonary tuberculosis. Ind J Tuberc. 1977;24:159–62. [Google Scholar]
- 18.Hidalgo JA, Vazquez JA. Candidiasis. emedicine [Internet]. Updated May 2012. http://emedicine.medscape.com/article/213853-overview. [Google Scholar]
- 19.Sanchez MS, Lloyd-Smith JO, Getz WM. Monitoring linked epidemics: the case of tuberculosis and HIV. PLoS One. 2010;5:8796. doi: 10.1371/journal.pone.0008796. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Revised National Tuberculosis Control Programme (RNTCP): Training Module for Medical Practitioners. New Delhi: Central TB Division, Directorate General of Health Services, Ministry of Health and Family Welfare. 2010:107p. [Google Scholar]
- 21.Pappas PG, Rex JH, Sobel JD, Filler SG, Dismukes WE, Walsh TJ. et al. Guidelines for Treatment of Candidiasis. Clin Infect Dis. 2004;38:161–89. doi: 10.1086/380796. [DOI] [PubMed] [Google Scholar]
- 22.Lovacheva OV, Kornienko I, Kul'ko AB. Efficiency of treatment of Candida-induced lower respiratory tract lesions in patients with pulmonary tuberculosis. Agrawal M. Candida in pulmonary tuberculosis. Probl Tuberk Bolezn Legk. J Postgrad Med. 2005 1982 Jan;28(1):14–6. 24–9. [PubMed] [Google Scholar]